Actin is an essential and highly conserved cytoskeleton protein that polymerizes into helical double- stranded filaments and powers a broad range of eukaryotic cell movements. The actin regulatory protein, cofilin, severs filaments and increases the number of ends from which subunits add and dissociate. Severing is critical for rapid filament growth at the leading edge, as well as subunit turnover and network remodeling. Modulation of actin filament bending and twisting elasticity has been linked to regulatory and contractile protein function, filament assembly dynamics, and overall cell motility. A quantitative molecular description of actin filament elasticity is therefore central for developing predictive physical models of cell mechanics and actin-based motility. Research efforts in this proposal focus on indentifying the molecular origins of actin filament elasticity and the mechanical basis of filament severing by cofilin. Two general hypotheses will be tested. The first is that the double-stranded, helical structure of actin filaments gives rise to a strong coupling of twisting and bending motions that dominates the filament elastic free energy at small deformations associated with normal cellular function. The second is that twist-bend coupling causes stress to accumulate locally at regions of mechanical and topological asymmetry, such as junctions of bare and cofilin-bound segments of partially-decorated filaments, thereby increasing the severing probability at these sites. We will integrate mathematical modeling and all-atom molecular dynamics simulations with experimental manipulation of single filaments to develop predictive molecular models of actin filament mechanics and test hypotheses formulated from biochemical and biophysical analysis of cofilin-actin interactions. We will develop mesoscopic actin and cofilactin filament models that capture key features, including subunit dimensions, interaction energies, helicity and the double stranded structure. Model filaments will be strained with external mechanical (buckling or torque) loads and the emergence of twist-bend coupling be assessed from out of plane deformations. Direct twisting manipulation of individual actin filaments will test predictions of actin filament elasticity made by the computational models. Evaluation of model filaments with different architectures (e.g. number of strands and helicity) will reveal the geometric origin of twist-bend coupling. The elastic free energy and twist shear density of model filaments will be determined to evaluate how twist-bend coupling contributes to stress accumulation and severing at boundaries of bare and cofilin-decorated filament segments. The proposed activities will provide an explicit link between the microscopic properties (filament radius, monomer dimensions, buried subunit interface area, lateral or longitudinal contacts), the global mechanical behavior (bending, twisting deformation and twist-bend coupling) of filaments and the biological function (e.g. severing activity) of essential regulatory proteins. General principles regarding the relation between helical biopolymer elasticity, structure and stability will emerge from this work. 1